DocumentCode
3197128
Title
Neural speech enhancement using dual extended Kalman filtering
Author
Nelson, Alex T. ; Wan, Eric A.
Author_Institution
Dept. of Electr. Eng., Oregon Graduate Inst., Portland, OR, USA
Volume
4
fYear
1997
fDate
9-12 Jun 1997
Firstpage
2171
Abstract
The removal of noise from speech signals has applications ranging from speech enhancement for cellular communications, to front ends for speech recognition systems. Spectral techniques are commonly used in these applications, but frequently result in audible distortion of the signal. A nonlinear time-domain method called dual extended Kalman filtering (DEKF) is presented that demonstrates significant advantages for removing nonstationary and colored noise from speech
Keywords
Kalman filters; feedforward neural nets; filtering theory; noise; nonlinear filters; speech enhancement; state estimation; cellular communications; colored noise; dual extended Kalman filtering; neural speech enhancement; nonlinear time-domain method; nonstationary noise; speech recognition systems; speech signals; Additive noise; Filtering; Kalman filters; Maximum likelihood estimation; Neural networks; Recursive estimation; Signal processing; Speech enhancement; Speech processing; State estimation;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks,1997., International Conference on
Conference_Location
Houston, TX
Print_ISBN
0-7803-4122-8
Type
conf
DOI
10.1109/ICNN.1997.614244
Filename
614244
Link To Document